ESG Thema #23 - Digital Transition Series - The Social Dimensions of Responsible AI Deployment

AI is a material source of economic opportunity, but its benefits will depend on responsible deployment: AI can support productivity, innovation and efficiency, yet weak governance of AI-related risks may undermine these gains. This leads to a strong case for investor engagement.

A governance gap continues to widen between AI commitments and implementation: Many companies now have ethical AI principles, but disclosure on practical controls remains limited and lags considerably behind AI diffusion.

Social risks arise across the AI value chain, with deployment emerging as a key investor focus: Developers shape model behaviour, but a large share of investors’ portfolio-level social risk sits with companies deploying AI across established workplace processes, products, services and consequential decisions.

AI is reshaping work in ways that extend beyond headcount effects: The main risks relate not only to displacement, but also to work redesign, weaker worker voice, algorithmic management and the erosion of formative early career tasks that build future human judgement and oversight capacity.

Downstream harms are becoming more visible and more material: AI-related discrimination in access to services, child safety failures, dual-use applications and community tensions linked to data centre expansion show how weak controls can create harm at both individual and systemic levels.

Investor stewardship should focus on evidence of a governed AI lifecycle: Investors should expect companies to demonstrate meaningful human oversight of AI, with ongoing monitoring and viable routes to remedy and adaptation as AI evolves.